Organizations such as the World Bank and the World Food Programme have been ahead of the curve, testing how different mobile methodologies compare to traditional F2F methods, or providing respondents recruited in-person with mobile phones for follow up studies. Others have used remote methodologies when in-person data collection was inadvisable, such as during the 2014-2015 West African Ebola outbreak, but have primarily relied on in-person methodologies where possible.

As the COVID-19 outbreak continues it is clear that for researchers, coronavirus may represent a tipping point in the adoption of remote research methods. Even with precautions in place, face-to-face interactions put both interviewers and respondents at higher risk of contracting a devastating virus, and yet there is a pressing need for data on both COVID-19 itself and the impact it will have across countries and sectors. The question has become not if researchers will need to transition to remote-based methodologies, but how we as a research community can make the most of this unanticipated shift to remote data collection.

Mobile-Based Research in Emerging Regions: Where We are Now

When GeoPoll began offering surveys via 2-way SMS in 2012, we encountered obstacles surrounding not only the buildout of our platform, but also a resistance to the idea that reliable, valid research could be conducted through remote surveys administered through mobile phones. Entering the industry as a mobile technology company, we did not recognize all the nuances and biases of gathering data through methodologies which had not yet been tested at scale.

We overcame questions surrounding the representativeness and quality of data collected through mobile by implementing strict quality control systems and running research experiments that expanded our knowledge of the strengths and weaknesses of various mobile modes. One study we conducted found that select-all-that-apply style questions have a lower number of responses in SMS than forced-choice questions, while another found that drop-off rates increase for SMS questionnaires as they get longer.

Studies by other organizations have examined response and completion rates, and the sample skew of various modes. Research on mobile samples has found that mobile phone owners tend to be more educated than general populations, and that the population reached varies depending on the mode uses. Studies have also looked at how to improve SMS opt-in rates, finding that higher incentives don’t lead to higher response rates, but that sending reminders can improve the participation of older and less educated respondents. The World Food Programme has shown that mobile surveys can accurately demonstrate food security trends at a lower cost than face-to-face interviews, and The World Bank has conducted various studies during health crisis or conflict which demonstrate the usefulness of mobile surveys especially during rapid-response situations.

The Future of Mobile-Based Research in Emerging Region

Although mobile-based research in countries in Africa, Asia, and Latin America has increased in the past few years, it has often supplemented in-person research. For fully-mobile studies such as those being conducted during COVID-19, Computer Assisted Telephone Interviewing, live voice calls conducted from a call center or remotely, has emerged as a favored mode as it allows for longer interviews than SMS, is interviewer-administered, and can reach illiterate populations.

While limited research is available that directly compares results from CATI studies to face-to-face research, studies have found CATI to have higher response rates than other mobile modes, and that CATI is better at reaching older populations than SMS or IVR, an automated voice-call method. During COVID-19, CATI is being used by multiple academic and governmental groups who are looking to balance the need for remote data collection with research requirements. GeoPoll itself has seen a huge increase in demand for CATI surveys, resulting in the completion of nearly 100,000 CATI interviews in the month of July alone.

In research, there is often a fear of the unknown, but the benefits of remote data collection – namely the relative speed, wide reach, and relatively low cost – often outweigh the methodological challenges. During a time when in-person research is impossible, researchers have the ability to test these methods at scale so they can be used not only during times of crisis, but incorporated as part of a viable research approach in all projects. Research is ever-evolving, and at a moment when many are being forced to experiment with new methods, we have the opportunity to document learnings that will benefit the industry even after the pandemic has subsided.

The post COVID-19: A Tipping Point for Remote Research appeared first on GeoPoll.

Computer Assisted Personal Interviewing

Computer Assisted Personal Interviewing is a face-to-face mode similar to pen and paper surveys. The difference lies in the technology used for record-keeping. In this mode, the interviewer reads the questionnaire off of a device screen (like a tablet) and records the answers electronically throughout the survey. Surveyors use CAPI software programs on mobile devices while conducting these face-to-face surveys, however, internet connectivity is not usually necessary. This makes data inputting and uploading a much simpler process and virtually eliminates the possibility of missing or damaged data files.

CAPI surveys have a higher associated cost and longer timeline than survey methods that can be completed remotely. However, CAPI surveys have the ability to reach every socioeconomic level in the desired survey area, even people living in extreme poverty or those without mobile access. CAPI surveys also often allow for longer surveys than remote methods of data collection, as the survey is conducted in-person rather than being self-administered.

SMS (Short Message Service)

Short message service, or text messaging, is a popular method of data collection in emerging markets where a higher percentage of the population has access to a mobile phone than the Internet. SMS surveying can also accomplish goals quickly without depleting budgets, as surveys are conducted remotely without the need for enumerators.

SMS surveys are sent to respondents one question at a time. All respondents are required to opt-in for participation after receiving the first message. The surveys are always self-guided and the answers self-reported through two-way messaging. The data collected through SMS surveying is largely quantitative, though some open-ended questions can be included. Surveys of this type are best when kept relatively short and each individual message must be under 160 characters—including answer choices. Despite the simplicity of SMS messages, SMS surveys can still include advanced features such as survey routing, skip logic, and question or answer choice randomization. It is important when evaluating a survey partner to ensure they have these capabilities, which GeoPoll’s platform supports.

SMS is a useful mode in countries with unreliable connectivity, as it does not require respondents to have internet connectivity, and messages can be received even if the phone is powered off or out of service range when the initial message is sent. Using SMS as a survey method ensures that any mobile phone owner can participate as a survey respondent, whether they have a smartphone, feature phone, or basic phone, which means SMS surveys are more representative than those conducted online in many emerging markets. SMS surveying can also accomplish goals quickly without depleting budgets, as surveys are conducted remotely without the need for enumerators.

CATI (Computer Assisted Telephone Interviewing)

CATI surveys are administered remotely through call center enumerators who interview survey participants. Compared to SMS surveys, live calling questionnaires can be longer and each question is not limited to 160 characters, which is beneficial because the resulting data can be more in-depth than in other methods. Interviewers read questions from a set script, and survey routing and skip logic can be incorporated into the script. This type of interview is more structured than a qualitative interview, in which answers to each question may lead the discussion in a certain direction.

CATI has several benefits: Longer questionnaires can be administered and more open-ended data can be collected. In addition, there is no reading required by the respondent for CATI surveys, making it a good mode of research collection for populations with low literacy rates.

When choosing CATI as a method, researchers must consider the effort and time involved in training call-center interviewers, and know who will be conducting the training. In emerging markets, interviewers may need to be fluent in multiple languages. GeoPoll trains all call center interviewers as part of our research services, and ensures that the interviewers used are based locally. We have established call centers in Kenya and Nigeria and have experience conducting CATI projects in many markets throughout Africa, Asia, and Latin America.

Data collected via CATI must also be recorded properly in order to be organized upon data delivery. There are a variety of software tools available for this purpose, and GeoPoll has a mobile application for CATI interviewers that facilitates interviewer flow and data collection on any smartphone or tablet. GeoPoll’s application is designed for use in emerging regions, and includes offline capabilities, allowing immediate access to survey results and interviewer metrics.

IVR

IVR surveys are another mode of remote survey data collection. This technique entails using pre-recorded voice bites that are initiated via phone calls. IVR survey participants self-report their responses by using their mobile device keypad to key-in responses.

Similar to CATI surveys, IVR is useful for reaching illiterate populations through voice calls—without the obstacle of training call center interviewers. Although not having to train interviewers saves time, an IVR project typically takes longer from beginning to end.

IVR surveys are commonly slower and less cost-effective than the majority of other similar modes of research. Lower levels of survey participation draw out the time it takes to reach completion on a particular project, which may increase overall costs. For these reasons, the questionnaires should be kept short in order to avoid response drop-off mid survey. Due to low participation rates, when using IVR companies must partner with a research company that has a large database of respondents to pull from. GeoPoll has a database of over 240 million mobile phone subscribers which allows us to administer IVR surveys that reach the desired sample size.

Mobile Web

Mobile web is a link-based survey that can be taken on a very basic mobile phone browser, such as those common throughout emerging markets. The mobile-web survey technique can be dispatched through a few different channels. The technique that GeoPoll uses most often combines SMS and web-based surveying. In order to collect data through this mode, an SMS text is sent to potential respondents. The users who receive survey invitations are directed to a webpage where they opt-in to complete the survey and self-report their answers.

The advantages of the mobile-web method surround increased capabilities. With mobile-web surveying, users can receive pictures and videos, surveys are not character-limited, matrix and other advanced question types can be used, and it is relatively inexpensive to conduct.

Mobile web’s reach within a country’s population depends on how the link is initially sent. If sent via SMS, those with basic mobile browsers can participate, however, surveys must be set up to be viewed correctly on these types of browsers. Mobile-web does always require that participants have Internet connectivity or a data plan, which means that most respondents through mobile-web are middle class or higher. The chart below represents mobile-web in comparison to the other modes that have been discussed.

Mobile Smartphone Application

Many market research companies connect with survey respondents through smartphone applications in established markets, yet in emerging markets, mobile application survey research has more limitations. Some of the challenges associated with mobile applications in these areas have to do with limited connectivity.

Mobile applications require users to have an Internet or data connection, a smartphone, and the user must have the survey application or download it as part of the recruitment process. Once the application is downloaded, surveys can be sent to potential respondents through the mobile application. Survey participants then self-report their answers and submit the surveys for review. Overall, remote mobile-based survey research through a smartphone application is an economical route that provides significant features for the questionnaire itself, yet the drawback is that in many emerging markets it is not as accessible for the masses as an SMS based survey would be.

It is important to note though that not every emerging market is the same. At GeoPoll, we have spent years studying how to best connect with survey respondents in emerging markets across the world. Picking the right mode for accomplishing your research goals is best done with the professionals. Contact GeoPoll today to talk about what research mode is best for your project.

The post Comparing Mobile Research Modes for Emerging Markets appeared first on GeoPoll.